[MPlayer-users] -vf ilpack

[D Richard Felker III]

On Sat, Mar 06, 2004 at 10:35:38PM +0200, Ville Saari wrote:
> On Fri, Mar 05, 2004 at 08:24:17PM -0500, D Richard Felker III wrote:
> 
> > If you think about it, it's actually NECESSARY that sampling be done
> > this way, since there's no way to tell reliably if a frame is
> > interlaced or progressive.
> 
> I thought that DVD players would simply use progressive sampling for
> progressively encoded mpeg and "ilpack" sampling for interlaced mpeg.

AFAIK there really isn't such a thing as "progressive" vs "interlaced"
in mpeg. You can use interlaced motion estimation or dct, or even
field-based encoding, but none of these is intended to reliably
represent whether the _content_ is interlaced or progressive. In fact
in some cases it might be optimal to encode interlaced frames without
any of the interlaced features (e.g. if there's almost no motion) and
sometimes even progressive frames with interlaced features (e.g. if
there's "combing" as part of the original progressive picture).

> Theoretically that SHOULD be a reliable way to tell, but in reality
> so many DVDs seem to be encoded by incompetent fools.

Agree totally about the incompetence part.

Whether it was _intended_ that you be able to tell if video is
progressive or interlaced from this info or not, you're right that in
practice it's totally impossible.

> Several PAL DVD are encoded so that progressive film content was
> converted to interlaced video, which was then encoded to DVD with
> opposite field dominance so that every frame is like the interlaced
> frames of telecined NTSC. I couldn't find a way to fix such video
> with mplayer's existing filter arsenal, so I wrote my own "phase shift"
> filter.

Uhg. Do you have any idea what this does to the quality/bitrate?? :(

> But some such DVDs are doubly blundered: Those interlaced frames are
> encoded as progressive mpeg! There is no way to acceptably fix such
> content because the fields will be partially blended together in the
> decoded picture. If I use my phase shift filter on that, the result
> is a weird combined ghosting and combing artifact: one field has future
> ghost and the other field has past ghost.

Applying the linear blend filter afterwards should clean it up enough
to be watchable without hurting the video quality too much. It will
halve the ghosting in each field and spread both ghosts out over both
fields so you don't see combing.

Rich